| 基于地理探测器与层次分析法的岩溶地面塌陷易发性评价——以重庆中梁山地区为例 |
| |
| 引用本文： | 王桂林,强壮,曹聪,陈瑶,郝晋渝.基于地理探测器与层次分析法的岩溶地面塌陷易发性评价——以重庆中梁山地区为例[J].中国岩溶,2022,41(1):79-87. |
| |
| 作者姓名： | 王桂林 强壮 曹聪 陈瑶 郝晋渝 |
| |
| 作者单位： | 1.重庆大学土木工程学院，重庆400045 |
| |
| 基金项目： | 重庆市技术创新与应用发展专项面上项目cstc2019jscx-msxmX0303重庆市规划与自然资源局科技项目KJ2019047国家重点研发计划课题2018YFC1505501 |
| |
| 摘 要： | 岩溶地面塌陷作为我国西南地区主要地质灾害类型之一，已成为影响该地区经济发展的重要因素，建立符合地区特征的岩溶地面塌陷易发性评价模型，可为塌陷的防治提供指导。文章以重庆市中梁山地区为研究区，以327个塌陷点为样本，基于GIS技术和地理探测器方法，对研究区三组样本点进行因子探测，定量化筛选出影响较大的评价因子，并采用层次分析法对岩溶地面塌陷易发性做出评价。结果显示:随着样本点数量变化，影响因子的解释度q值排序存在差异，然而三组数据中，各因子对岩溶塌陷的贡献大小排序始终是地层类型、地层富水性、距隧道距离、高程和坡度；基于地理探测器—层次分析法相结合的岩溶塌陷易发性评价结果预测精度达89.88%，高易发区主要在岩溶槽谷区嘉陵江组和大冶组地层分布地段。
|
| 关 键 词： | 岩溶塌陷 易发性 地理探测器 层次分析法 |
| 收稿时间： | 2020-05-30 |
Evaluation of susceptibility to karst collapse based on the geodetector and analytic hierarchy method: An example of the Zhongliangshan area in Chongqing |
| |
| Abstract: | Karst collapse is a process in which the surface collapses suddenly under the action of various factors in the karst distribution area. As one of the main types of geological disasters in Southwest China， karst collapse mainly damages roads， railways， buildings and surface water， influences the use of agricultural water and land， and causes casualties and property losses. Due to its covertness， suddenness and uncertainty， karst collapse has become an important factor affecting the regional economic development. Therefore， establishing an evaluation model of karst collapse susceptibility that conforms to regional characteristics is of great significance for local the planning of land use and collapse prevention.This research takes the Zhongliangshan area of Chongqing as the study area， and samples 327 collapse points from the field investigation. Based on the factors that may induce the karst collapse， 13 potential impact factors of 6 classes for karst collapse have been preliminarily determined， namely， the topography （elevation， slope， slope direction， surface curvature， section curvature， slope position， and surface roughness）， the geological structure （distance from fault）， strata， hydrogeology conditions （formation rich in water， and terrain humidity index）， human engineering activities （distance from tunnel）， overburden characteristics （soil thickness） and so on， and a geospatial database of the study area has been established by using GIS to process the original data. Given the influence of the sample number in non-collapse points on the selection of impact factors， this study uses three groups of number ratios in different collapse points to analyze the explanatory power （q-statistic in Geodetector） of each factor in the karst collapse area. In order to avoid the fact that the establishment of the pair comparison matrix in the analytic hierarchy process is too tedious and inefficient， resulting from the excessive impact factors， the factor detection has been carried out in three groups of sample points in the study area， and the evaluation factors with greater influence on collapse have been selected quantitatively by the size of q value， based on GIS technology and geographic detector method. According to the principle of analytic hierarchy process and the screening results of subsidence impact factors， the evaluation system of karst collapse susceptibility in the study area has been established by taking the karst collapse susceptibility as the target layer. In order to accurately reflect the important difference among factors and reduce the influence of human experience factors， a pair comparison matrix has been established based on the collapse distribution， impact factor analysis results and q value results of geographical detector. The susceptibility to karst collapse has been evaluated by using the analytic hierarchy process. With the help of the GIS spatial analysis module， the evaluation results have been assigned to grid units， and then the zoning map of the karst collapse susceptibility in the study area is obtained. The results show that as the sample amount changes， there is a degree difference in the importance of impact factors. However， among the three sets of data， strata， formation rich in water， distance from tunnel， elevation and slope are always the factors that have the largest impact on karst collapse. The use of geographic detectors to filter factors can avoid the influence of irrelevant factors， and the prediction accuracy （89.88%） conducted by analytic hierarchy process to zone the karst collapse susceptibility has been significantly improved. The areas with higher probability of collapse mainly distribute in the Jialingjiang formation and Daye formation in the karst trough area. |
| |
| Keywords: | |
|
| 点击此处可从《中国岩溶》浏览原始摘要信息 |
| 点击此处可从《中国岩溶》下载免费的PDF全文 |
